CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation

TL;DR

CoSMix introduces a novel sample mixing approach for unsupervised domain adaptation in 3D LiDAR segmentation, leveraging a dual-branch network to improve model generalization across different sensors and environments.

Contribution

This is the first UDA method for point cloud segmentation based on sample mixing, combining semantic information from source labels and target pseudo-labels.

Findings

Outperforms state-of-the-art methods on large-scale datasets

Effective in mitigating domain shift in 3D LiDAR segmentation

Demonstrates significant improvements in real-world autonomous driving scenarios

Abstract

3D LiDAR semantic segmentation is fundamental for autonomous driving. Several Unsupervised Domain Adaptation (UDA) methods for point cloud data have been recently proposed to improve model generalization for different sensors and environments. Researchers working on UDA problems in the image domain have shown that sample mixing can mitigate domain shift. We propose a new approach of sample mixing for point cloud UDA, namely Compositional Semantic Mix (CoSMix), the first UDA approach for point cloud segmentation based on sample mixing. CoSMix consists of a two-branch symmetric network that can process labelled synthetic data (source) and real-world unlabelled point clouds (target) concurrently. Each branch operates on one domain by mixing selected pieces of data from the other one, and by using the semantic information derived from source labels and target pseudo-labels. We evaluate CoSMix on two large-scale datasets, showing that it outperforms state-of-the-art methods by a large margin. Our code is available at https://github.com/saltoricristiano/cosmix-uda.